Mind The Gap: eResearch in Australia
Paul Davis, Gabrielle Bright
Victorian eResearch Strategic Initiative, Australia
Abstract
VeRSI - the Victorian eResearch Strategic Initiative - was funded in 2006 by the Victorian State
Government. This paper will describe how the VeRSI program is providing a cohesive and
coordinated approach to accelerating the uptake of eResearch by Victorian researchers. This is the
first initiative of its kind in Australia.
VeRSI will deliver research leadership by harnessing enabling technology in a way that fosters a
productive, collaborative research environment, and by developing key examples of how eResearch
can enhance research outcomes. This paper will describe the five complementary projects that will
establish essential infrastructure, develop resources for collaborative working and support exemplar
use-cases and the development of applications in the Life Sciences.
VeRSI, like similar eResearch activities elsewhere, is acutely aware of the IT skills shortage
developing in Australia. This exacerbates the need to develop skills and expertise in eResearch.
This paper will highlight the need to bridge the skills and expertise 'gap' between research
disciplines and technology implementations.
1. Challenges to research in the past
In the nineties, research communities in science
and engineering began to tackle the so-called
“Grand Challenges”; fundamental problems
with broad application. The Grand Challenges
were so large and complex that they could only
be addressed by large teams using supercomputers and would take decades to solve.
Studies in computational fluid dynamics would
enable the optimal design of vehicles, provide
more accurate weather forecasts and improve
efficiency in the oil industry. A computational
approach to modelling the electronic structure
of matter would deliver new chemical catalysts,
drugs to boost the immune systems and novel
superconductors. The physics underpinning
fusion technology would lead to the energy
sources of the future and bioinformatics and
computational chemistry would provide the
fundamental understanding of molecular
systems necessary to cure cancer, diabetes,
cardiovascular ailments and more.
A watershed in computer science provided the
algorithmic tools needed for such endeavour
and
technology
delivered
phenomenal
improvements in the power and speed of high
performance computer systems. One by one the
Grand Challenges were reduced to large but
routine computations and even more exigent
challenges filled their place; but the nature of
the problems and, indeed, of their solutions had
changed dramatically.
Today’s Grand Challenges are global and
involve health, climate change and global
warming, the environment, predicting and living
with extreme seismic activities and tsunamis,
disaster reduction and security and the spread
and containment of infectious disease. These are
fundamental issues that effect life on the planet,
not just the profitability of the manufacturing
industry. They demand profound understanding
of exceedingly complex systems and they
simply cannot be solved by teams constrained to
a department, institution or organisation.
Figure 1: The change in science focus with
technology.
Figure 1 illustrates the continuum of change in
research with the influence of ICT and
transition to meet new challenges over time.
2. System Science
Science has become a team sport and we have
entered the era of “System Science” - the
integration of diverse sources of knowledge
about the constituent parts of a complex system
with the goal of obtaining an understanding of
the system's properties as a whole. System
Science is characterised by global collaboration,
shared resources, common goals and multidisciplinary problem solving. Ian Foster, one of
the recognised “fathers of the grid” provides the
following commentary:
“System-level science integrates not only
different disciplines but also, typically, software
systems, data, computing resources, and people.
System-level science is usually a team pursuit.
Data comes from different sources, different
groups develop component models, team
members provide specialised expertise, and the
often substantial computing and data resources
required for success are themselves diverse and
distributed. Thus, system-level science itself
requires the creation of yet another sort of
system that may combine large numbers of both
physical and human components.”i
An example of System Science is a virtual
organisation (VO) for the study of
ecoinformatics: an online, collaborative and
shared resource for managing ecosystem and
environmental data and information products.
The ecoinformatics VO becomes the focal point
for research collaborators and policy
participants: a common, virtual platform that
provides a consolidated data and knowledge
infrastructure to support research design and
delivery, and greatly improved collaboration
across research teams in a shared environment.
It brings together, as a base set, data and
methods from ecology, environmental science,
land care, water management, climate studies,
agriculture, animal husbandry, mathematics and
computer science and advanced ICT.
Treating ecoinformatics as a System Science
will lead to innovative methods for the
integration, curation and management, mining,
interrogation, visualisation and modelling of
multi-scaled (local, regional, national and
global) data within the shared platform, thus
delivering the best available and most costeffective research outputs to the natural resource
and primary industry sectors.
The growing awareness that researchers must
collaborate, that resources have to be shared,
that information and communication technology
has to be better integrated into the research
environment and that these characteristics apply
to all research activities not just science, has
resulted in the genesis of eResearch.
3 eResearch in Australia
Dr Mike Sargent, Chairman of the Australian eResearch Coordinating Committee (eRCC),
Australia’s equivalent of Sir John Taylor, wrote
in his report to the Government:
“On a global scale, the level of investment in
ICT enablement of research in the United
Kingdom, United States of America, Europe and
Canada is beginning to pay dividends. In the
United Kingdom, £250M has been invested in
the e-Science programme over the past 5 years
and the programme has drawn further
significant investment from corporations such
as IBM, Microsoft and Oracle. In the Asia
Pacific region, the governments of Japan,
Korea, Taiwan and China are strongly
committed to finance their e-Research
frameworks. Countries investing in e-Research
capabilities are increasingly participating in
collaborative research into the grand
challenges.”ii
Australian has been slow to adopt eResearch,
however, this tardiness has been an advantage in
a country with merely twenty-one million
people and fewer than eighty thousand
researchersiii. Australia has been able to watch
and learn from the development of eResearch
activities in the US, Europe, UK, Japan, Korea,
China, Singapore and Latin America and
develop a philosophy of “adapt – adopt – and
contribute” based on successes elsewhere.
In particular, Australia has modelled its
eResearch development on that of the UK. A
large number of delegations have explored
every aspect of UK eScience over the last five
years and several key emissaries, for instance,
Sir John Taylor, Dr A. Hay, Professors A.
Trefethan, C. Gobles, and D. DeRoure, have
visited Australia to advise on eResearch policy
and procedures.
VeRSI, the Victorian eResearch Strategic
Initiative, exists on a much smaller scale than
the UK wide eScience Program. Nevertheless,
the principles of awareness as the first step;
working with the research community to
establish trust; developing solutions for selected
research groups that facilitate collaboration and
tangible examples of the benefits of
Figure 2: Science value chain
collaborative research and shared resources,
and; providing key enabling infrastructure, are
the same.
The Australian Department of Education,
Science and Technology (DEST)iv proclaims:
“The research sector worldwide is experiencing
enormous change driven by advances in
information and communications technology
(ICT). Research is increasingly characterised
by national and international multi-disciplinary
collaboration and most OECD countries and
APEC members are investing heavily in those
capabilities and the associated coordinating
mechanisms.
The term ‘e-Research’ encapsulates research
activities that use a spectrum of advanced ICT
capabilities and embraces new research
methodologies emerging from increasing access
to:
•
Broadband communications networks,
research instruments and facilities, sensor
networks and data repositories;
•
Software and infrastructure services
that
enable
secure
connectivity
and
interoperability;
•
Application tools that encompass
discipline-specific tools and interaction tools.”
Apart from the Australian Government’s
insistence, there are several compelling reasons
why eResearch is becoming the new paradigm.
Robert Kelley (Carnegie Mellon University)
claims that, in 1986, 75% of the knowledge one
needed to do one’s job was stored in the mind.
In 1997 this had fallen to 15% to 20%v;
knowledge has become an on-line commodity.
Howard Garner (Harvard University) amplified
this message when he stated, “Knowledge does
not stop at my skin: it includes my computer and
its databases and my network of associates”vi.
eResearch provides the infrastructure and
methodologies for accessing and reusing on-line
knowledge.
Figure 2, above, is a schematic of the science
value chain illustrating that eResearch, by virtue
of the collaborative and shared dogma, provides
a seamless communications path for ideas, data
and knowledge from the fundamental research
carried out (mostly) in universities through to
industry.
Two years ago DEST commissioned the
Australian eResearch Coordinating Committee
to undertake a comprehensive review of
eResearch and recommend how Australia,
cognisant of experiences elsewhere, could
coordinate a National eResearch initiative. One
of the recommendations of the Committee was
the establishment of an eResearch Centre
consisting of a coordinating body and six statebased nodes to facilitate the transfer of
eResearch methodologies to the research
community.vii
4
Challenges
Australia
to eResearch
in
eResearch initiatives, such as VeRSI, are not
without their challenges; several key issues in
the adoption of eResearch have yet to be
addressed. These include the reward system for
academics, which is not geared to adequately
measure
outputs
other
than
papers.
Encouragement of open-source software, online publication and sharing of experimental
data and derived results all test the ability of the
research sector to measure real research outputs
and reward researchers accordingly. This is
particularly important as the research
community is encouraged to move from
competitive to collaborative endeavours.
Before researchers will adopt new methods
there has to be a clear, unambiguous and
demonstrable benefit. In chemistry, the reaction
coordinate is an abstract one-dimensional
coordinate system that represents progress along
a reaction pathway (Figure 3). Reactants are
mixed, energy is added (the activation energy)
and the reaction proceeds to the formation of the
wanted products. If insufficient energy (less
than the activation energy) is added the reaction
cannot proceed. By analogy, for a researcher to
move to a new paradigm the “product”
environment must be better than the starting
point AND the “activation energy” (learning,
re-training, disruption to research activity, costs
of change etc) has to be very small. VeRSI’s
role is to describe and demonstrate ways to
adopt eResearch that lead to better research
environments and require little or no energy
input from the researchers. The hidden catch is
in the subjective term “better”; in this context
“better” is determined by the individual
researcher.
undertaking research training in eResearch
capabilities at the PhD level. It was proposed
that this be supplemented with the introduction,
by institutions, of formal incentives, recognition
and reward mechanisms for the skilled
eResearch professionals who support and
provide expertise to researchers, and to
encourage academics to invest time and
expertise to develop cross-disciplinary projects
and courses to train a new generation of
eResearchers.
The press (iTWire) report that:
“Salaries in the Information, Communication
and Technology sector were up by 12% across
the board for the six months to December
2006…”
and, “…demand for application developers,
data management professionals, and business
analysts was continuing unabated with this
trend expected to go on strongly in 2007.
“The market has been, and remains, hot with
demand outstripping supply,” Andy Cross,
Ambition technology managing director said.
“Salaries have jumped 12% and the technology
recruitment market can’t bridge the demand
gap. That means the war for talent will continue
in 2007. This will be the key challenge in our
sector for next year.”” viii
With this level of competition and the
uncompetitive salaries and conditions offered
by the university sector the recruitment
challenge will continue to be the major gap
between the ideals of eResearch and its
realisation.
5 eResearch in Victoria
Figure 3: Chemical reaction pathway – an
analogy for change
The most significant challenge, however, is the
paucity of skilled ICT personnel, particularly
those with multidisciplinary expertise, and the
difficulties of recruiting staff who can bridge
the void between research disciplines and
technology developers. This was clearly
identified by the eRCC who recommended the
establishment of an extensive training program
involving a five-year programme of one-year
eResearch Honours Scholarships and a five-year
programme
of
three-year
eResearch
Postgraduate Scholarships to support students
The Victorian State Government, through
Multimedia Victoria, decided to meet these
challenges and in 2006 announced funding for
the Victorian eResearch Strategic Initiative
(VeRSI) as part of its Healthy Futures: Life
Science Statementix. The government funding
will be supplemented by equal amounts of inkind support from the partners. The clear
intention is that VeRSI become the Victorian
node of the national eResearch Centre.
The VeRSI Program is an unincorporated joint
venture. The initial members of the Consortium
are Melbourne University and Monash
University
along
with
the
Victorian
Government. VeRSI is based on collaborative
and inclusive model and will openly involve
research groups, skills and services from other
universities, research organisations, government
departments and service providers such as the
Australian Synchrotron, Victorian Partnership
for Advanced Computing (VPAC) and
Victorian Education and Research Network
(VERNet).
The funded activity will proceed from October
2006 to September 2010. During this period the
VeRSI
Program
will
undertake
five
complementary projects to provide support and
services to researchers, undertake an extensive
outreach and awareness raising exercise, deliver
a coordinated program of skills development
and assist in the coordination of research
development and deployment.
The five projects are broken into three
categories: enabling projects, a demonstration
project and capability projects.
• Security & Access: Security and access
includes the means to identify and authenticate
users, to provide them with access to shared
resources and to provide assurances that
information stored in shared repositories is safe
and secure. The challenge is to balance the
security and access needs of the community of
interest with those of the host institution(s).
VeRSI are developing a standards-based
security and access policy that is compatible
with existing systems and implementing the
supporting ICT platform so that authenticated
access is possible from anywhere.
• Storage Systems: Data storage is a high
demand commodity and most university
systems are over taxed and designed specifically
to meet the immediate needs of the campus.
This leaves little scope for experimentation with
storage standards and protocols. The UK
eScience program successfully addressed this
issue by providing limited, production quality
resources, particularly data storage and
management, as a foundation on which to build
trust and deliver services with minimal impact
on University security systems and firewalls.
VeRSI will follow this example.
5.2 Demonstration Project
Figure 4: Relationships between the VeRSI
projects
5.1 Enabling Projects
The enabling projects will create awareness and
establish essential infrastructure such as an
access and security framework and distributed,
federated storage facilities. There are three
enabling projects:
• Communications
&
Support:
Communications are a critical component in the
development of an eResearch framework
Informing researchers, explaining the benefits
of eResearch methods and assisting them to
assess their requirements and general
engagement are all key, on-going activities that
form
the
Awareness,
Outreach,
Communications and Support project.
The demonstration project is the design and
construction of a prototype Virtual Beam Line
(VBL) to demonstrate remote working with the
High-throughput
Protein
Crystallography
beamline at the Australian Synchrotron. The
VBL will allow for collaborative working as
part of grid-enabling the Australian Synchrotron
and synchrotron user community. The VBL will
provide a multimedia nexus through which
users can collaborate using voice, video and
shared applications, undertake occupational
health and safety, radiation and beamline
training, use a variety of tools to transfer data to
external storage and computing resources and
remotely monitor and mentor their experiments.
The VBL design will be modular, easily
replicated and meet the need of the synchrotron
community. The design will be promoted and
freely available to interested parties at other
beamlines.
5.3 Capability Project
The capability project is a series of exemplar
use-cases and applications for researchers in the
Life Sciences, capitalising on Australia’s
existing research strengths. These projects will
deliver production quality tools to the specific
research groups and serve as tangible examples
to the research community of how advanced
ICT that is responsive to researchers’ needs can
enhance the research environment. There are
eight use-cases:
• Genomic dataset mining: The function of
genes is dependent on their structure but, due to
the size of genomic data sets, mining of the
gene structure is an arduous task. Further
insights are developing from the merging of
publicly available data with proprietary
information. Innovative methods for collection,
curation, visualisation, integration and mining
of these data will deliver a major advantage to
the Victorian biotechnology and biomedical
research sectors.
VeRSI is supporting this research by designing
the general architecture of software modules for
data mining, assisting the implementation of
novel data mining techniques and building a
user interface to the data sets.
• Neurosciences & biomedical imaging:
The Neuroimaging Laboratory at the Howard
Florey Institute is using human and animal brain
imaging techniques to investigate clinical and
basic neuroscience research questions. These
include how various functions change with
variations in hormone levels, and whether
neurobiological changes can be observed prior
to the onset of physical symptoms in sufferers
of diseases such as Huntington's or Alzheimer's
disease.
VeRSI is supplying crucial integration and
networking capabilities, in addition to building
a metadata explorer model, a metadata editor for
neuroimaging data sets, a user interface to the
data and a data authentication model.
• Australian Mouse Brain Map: The
Mouse Brain Map Consortium is building a
national atlas of brain data, from MRI, histology
and immuno-histochemistry, for comparison
between diseased mouse brains and control
brains, studying changes in brain anatomy and
function.
VeRSI is building a repository of mouse brain
map images and information from institutes
within the consortium, designing the general
mouse brain map database architecture along
with dataflow, specimen tracking and security
models and building an interface to the
repository for upload, management, searching
and download of data.
• Uro-oncology Informatics Grid: Around
fifty percent of Australian men experience some
type of prostate problem during their lifetime.
Prostate cancer is one of the most common
forms of cancer in men. The Centre for
Urological Research aims to provide better
diagnosis and treatments for prostate cancer and
benign prostate disease. A biorepository of
prostate, kidney and bladder specimens will
allow researchers to share resources with
existing national and international repositories.
VeRSI is designing and developing an
informatics grid, ensuring the informatics grid is
interoperable between research centres and
national and international tissue banks, and
organising lab data, urology data, and pathology
data.
• Ambulatory motion studies: New
technology allows the easy collection and
examination of numerous gait parameters, over
many trials, for large groups of people within a
laboratory or field setting. The data provides
clinically valuable information about when gait
changes occur, leading to a improved
understanding falls and serious injury in older
adults.
VeRSI is improving the management,
distribution and access of gait motion study data
by designing and building the data architecture
for storage of gait study data and building a
security model for access to data sets.
• Workflows and laboratory automation
for metabolomics: The fundamental challenge
of the post-genomic era is to utilise the
information generated by genome sequencing
projects in conjunction with high-throughput
profiling technologies to understand cellular
functions on the molecular level in a
comprehensive
and
integrated
manner.
Metabolomics is an emerging tool increasingly
used to identify new protein functions and to
model the whole cell metabolism.
VeRSI is developing workflows and data
management models to automate metabolomics
research at the Bio21 Molecular Science and
Biotechnology Institute, designing a data
storage model for metabolomics data to capture
the entire data flow and designing and
implementing a web based user interface to
enable data management and provide access to
shared data.
• Distributed
radiotherapy
system:
Computation for radiotherapy patient treatment
planning requires high performance computing
(HPC) resources, but it is preferable for patients
to be treated as close to home as possible.
Distributed treatment planning allows for
planning to occur at HPC resources in
metropolitan areas after which it can be
distributed to regional clinics.
This use-case will provide a connection between
Peter MacCallum Cancer Centre and satellite
centres, via VERNet (Victorian Education and
Research
Network),
and
develop
an
authenticated portal interface to treatment
planning, imaging and clinical datasets.
• Laboratory data management: Australian
scientists need to build a critical mass and
develop workflow technologies and practices
that link geographically dispersed groups. This
will enhance collaboration in areas of niche
strength such as agrifoods biotechnology, stem
cell research, synchrotron-based research and
advanced clinical trials. It can be achieved by
combining
internet-based
collaborative
environments and grid computing, collectively
known as eScience, which in turn will underpin
the enhanced data handling and mining
methods.
This use-case will develop workflow
technologies
and
practices
that
link
geographically disperse groups and enhance
collaboration in areas of niche strength such as
agrifoods biotechnology, stem cell research,
synchrotron-based research and advanced
clinical trials
5.4 VeRSI Outcomes
VeRSI activities will provide support and
services to researchers, undertake an extensive
outreach and awareness raising exercise, deliver
a coordinated program of skills development
and assist in the coordination of research
development and deployment.
The outputs from the project groups will be
knowledge and expertise, designs and
technology solutions and advanced open source
software. All of these commodities have value
that will enhance the quality of research, bring
timelier
research
outcomes,
catalyse
international collaborations, and provide
opportunities to Victorian industry.
VeRSI will deliver research leadership by
harnessing enabling technology in a way that
delivers a productive, collaborative research
environment and by developing key examples
of how eResearch can enhance research
outcomes. This approach will catalyse the
widespread adoption and uptake of eResearch
and deliver the promise of faster and more
exhaustive research activities leading to
improved commercialisation opportunities and
an elevation in the status of the State’s academic
and research institutions. It will add value to the
Australian Synchrotron, improve collaboration
in the Life Sciences and reinforce Victoria’s
position as a knowledge-based economy.
6 Conclusion
The adoption of eResearch is a cultural change
process which, paradoxically, will have
succeeded when the “e” is no longer necessary
and the “research method” encompasses all the
tenets of shared, collaborative working enabled
by advanced information and communication
technologies. But this will take time. We can
learn the sociological, physiological and
technological lessons of the UK, Europe and the
US all of whom lead Australia in the adoption
of eResearch but this will have little effect on
the speed with which the culture of research
changes. The “e” will be part of the promise for
a time to come.
i
I. Foster, C. Kesselman, " Scaling SystemLevel Science: Scientific Exploration and IT
Implications.” Computer Magazine, vol. 39, no.
11, 2006, pp. 31-39.
ii
eResearch Coordinating Committee, Final
Report of the e-Research Coordinating
Committee,5
DEST,
http://www.dest.gov.au/sectors/research_sector/
publications_resources/profiles/e_research_strat
_imp_framework.htm, 2004)
iii
Research and Experimental Development, All
Sector Summary, 8112.0, Australian Bureau of
Statistics,
iv
http://www.dest.gov.au/sectors/research_sector/
policies_issues_reviews/key_issues/e_research_
consult/default.htm
v
R.E. Kelley, How to Be a Star at Work: 9
Breakthrough Strategies You Need to Succeed,
Three Rivers Press 1998,p77.
vi
H. Gardner,
Intelligence:
Multiple
Perspective, Wadsworth Publishing 1995
vii
http://www.dest.gov.au/sectors/research_sector/
publications_resources/profiles/e_research_strat
_imp_framework.htm
viii
http://www.itwire.com.au/content/view/7503/50
/
ix
http://www.business.vic.gov.au/BUSVIC/STAN
DARD//pc=PC_61353.html